| Table
1Representative FIFO
devices1 |
| |
Major
characteristics &
special features |
Depth
range |
Width
range
(bits) |
Maximum
frequency
(MHz) |
Retransmit
feature |
Status
flags2 |
First-word
fall-through |
Special
noise-control
features |
Cascading
methods |
Approximate
price per
kilobyte3 |
Second
sources |
| Cypress |
Small, asynchronous/
unidirectional, parity generate and check |
64 bytes |
4/5/8/9 |
20 |
No |
HF, AFE |
No |
Output slew-rate control |
Serial and parallel |
$80 |
AMD, IDT |
| |
Asynchronous/
unidirectional, parity generate and check |
256 bytes to 32 kbytes |
9 |
100 |
Read pointer resets to 0 |
EF, FF, HF on 42×/3×/6×, add AE,
AF on 47× |
No |
Output slew-rate control |
Serial and parallel, single token |
$25 (256 bytes)/
$0.75 (32 kbytes) |
AMD, IDT, TI, Quality Semiconductor |
| |
Clocked/unidirectional, parity
generate and check |
512 bytes to 4 kbytes |
9/18 |
83 |
No |
EF, FF, HF on 46×, add AE, AF on
47× |
No |
Output slew-rate control |
Serial and parallel, single token |
$15 (512 bytes)/
$5 (4 kbytes) |
No |
| |
Synchronous/
unidirectional, parity generate and check |
64 bytes to 8 kbytes |
9/18 |
100 |
On ×18s read pointer resets to 0 |
EF, FF, AE, AF, add HF on ×18s |
No |
Output slew-rate control |
Serial and parallel on ×18s, no on
×9s |
$75 (64 bytes)/
$1.90 (8 kbytes) |
IDT, TI, Quality, Semiconductor,
Sharp |
| |
Deep synchronous/
unidirectional, parity generate and check |
8 to 32 kbytes |
9/18 |
100 |
On ×18s read pointer resets to 0 |
EF, FF, AE, AF, add HF on ×18's |
No, but only 2.5 cycles latency |
Output slew-rate control |
Serial and parallel on ×18s (two
token), no on ×9s |
$1.30 (16 kbytes)/
$0.85 (32 kbytes) |
No |
| |
Double synchronous/
unidirectional, parity generate and check |
256 bytes to 8 kbytes |
9 |
100 |
No |
EF, FF, AE, AF, HF |
No |
Output slew-rate control |
Parallel support only |
$30 (256 bytes)/
$3.25 (8 kbytes) |
IDT |
| Integrated Device Technology |
Synchronous unidirectional and
bidirectional, mailbox-bypass register, byte
swapping, dynamic bus matching, parity generate
and check on 64-bit depth only |
64 bytes to 1 kbyte |
36 |
67 |
No |
FF, AF, AE, EF |
Yes |
Input blanking circuitry
desensitizes input noise |
Parallel and serial using first-word
fall-through with no need for external logic,
gate similar flags from different FIFOs |
$2 |
Some |
| |
Synchronous, bidirectional |
256 and 512 bytes |
18 |
50 |
Pointer resets to 0, extra counter
bit prevents pointer from getting ahead of write |
FF, HF, AF, AE, EF |
No |
Input blanking circuitry
desensitizes input noise |
No |
$1 |
No |
| |
Supersynchronous |
8 and 16 kbytes |
9 or 18 |
100 |
Pointer resets to 0, extra counter
bit prevents pointer from getting ahead of write |
FF, HF, AF, AE, EF |
Yes |
Input blanking circuitry
desensitizes input noise |
Parallel and serial using first-word
fall-through with no need for external logic,
gate similar flags from different FIFOs |
$0.76 |
No |
| |
Synchronous triple port, dynamic bus
matching |
64 bytes to 1 kbyte |
9/18/36 |
67 |
No |
FF, AF, AE, EF |
Yes |
Input blanking circuitry
desensitizes input noise |
Parallel and serial using first-word
fall-through with no need for external logic,
gate similar flags from different FIFOs |
$2 |
No |
| Motorola |
Synchronous, unidirectional,
dual-ported CMOS line buffer with a single clock |
8 kbytes |
8 |
50 |
Programmable read and write pointers |
NA |
No |
Unnecessary |
Expansion to eight chips by
assigning unique 3-bit value to each
devices control register |
$0.60 |
No |
| NEC |
Asynchronous line buffer,
unidirectional |
5 kbytes and 10 bytes |
8/16 |
40 |
Read pointer resets to 0 |
None |
No |
None |
Serial only using single-wire |
$3 per unit |
No |
| Sharp |
Asynchronous, unidirectional |
512 bytes to 8 kbytes |
9 |
100 |
Read pointer resets to 0 |
FF, EF, HF |
Yes |
Output slew-rate control |
Serial and parallel, single-wire
token-passing, use external logic to derive
composite flags |
$3.10 |
Industry standard |
| |
Synchronous, unidirectional |
512 bytes to 4 kbytes |
18 |
50 |
Programmable read pointer |
FF, EF, HF, AF, AE, HF synchronized
to either clock (selectable) |
No |
Output slew-rate control |
Default is two-wire token passing,
with optional pipelined mode, uses second
read/write enables and delayed flag |
$7.20 |
Quality Semiconductor, TI, IDT |
| |
Synchronous, unidirectional,
mailbox-bypass register, byte swapping, dynamic
bus matching, parity generate and check |
1 kbyte |
36 |
50 |
Read pointer resets to 0 |
FF, EF, HF, AF, AE, HF synchronized
to either clock (selectable) |
No |
Output slew-rate control |
Pipelined, cross-connecting flags
and enables provided for interlocked width
expansion |
$8.20 |
Quality Semiconductor |
| |
Synchronous, bidirectional,
mailbox-bypass register, byte swapping, dynamic
bus matching (512 and larger), parity generate
and check |
256 bytes to 1 kbyte |
36 |
55 |
Read pointer resets to 0 |
FF, EF, HF, AF, AE, HF synchronized
to either clock (selectable) |
No |
Output slew-rate control and series
resistor |
None |
$12.80 |
Quality Semiconductor |
| Synergy |
Single-clock, synchronous, very high
speed, parity generate and check |
64 bytes |
18 |
200 |
Yes |
FF, HF, EF, overflow, underflow |
No |
Proprietary bipolar ECL technology
plus voltage and temperature compensation |
None |
$1536 to $2048 |
No |
| Texas Instruments4 |
ACT, clocked, strobed, asynchronous,
synchronous/
unidirectional, bidirectional |
256 bytes to 16 kbytes |
1/4/5/8/9 |
67 |
Some |
FF, AF, HF, AE, EF |
All clocked devices |
TI's output-edge control |
Depth and width, FF and EF logically
AND connected |
$1 to $154 |
Some |
| |
ABT, clocked, strobed, bidirectional |
512 bytes |
18 |
80 |
No |
FF, AF, AE, EF |
All clocked devices |
TI's output-edge control |
Depth and width, FF and EF logically
AND connected |
$61 to $154 |
No |
| |
ACT, clocked, strobed,
unidirectional |
64 bytes to 1 kbyte |
18 |
67 |
No |
FF, AF, HF, AE, EF |
All clocked devices |
TI's output-edge control |
Depth and width, FF and EF logically
AND connected |
$2 to $205 |
No |
| |
ALV, clocked, strobed,
unidirectional |
64 to 512 bytes |
18 |
50 |
No |
FF, AF, HF, AE, EF |
All clocked devices |
TI's output-edge control |
Depth and width, FF and EF logically
AND connected |
$5 to $154 |
No |
| |
ABT, clocked, unidirectional,
bidirectional, mailbox-bypass register, byte
swapping, dynamic bus matching, parity generate
and check |
64 bytes |
32/36 |
67 |
No |
FF, AF, AE, EF |
All clocked devices |
TI's output-edge control |
Depth and width, FF and EF logically
AND connected |
$20.50 to $51 |
Some |
| |
ACT, clocked, unidirectional,
bidirectional, mailbox-bypass register |
256 bytes to 2 kbytes |
32/36 |
67 |
Yes |
FF, AF, AE, EF |
All clocked devices |
TI's output-edge control |
Depth and width, FF and EF logically
AND connected |
$5 to $51 |
Some |
|